DESCRIPTION (Applicant's Abstract):Cigarette smoking is an enormous public health problem. Although our understanding of the consequences of smoking is relatively advanced, our knowledge of the determinants of smoking cessation- and its frequent concomitant relapse-is far less complete. Enduring smoking cessation remains an elusive goal for most heavy smokers. The extraordinary addictiveness of nicotine is certainly a prominent cause of these difficulties, but cannot of itself explain the marked inter-individual differences in smoking cessation and relapse. The overarching goal of this proposal is to delineate etiological factors involved in smoking cessation and relapse. To accomplish this aim, we will: (I) obtain smoking history data on 40,000 individuals (20,000 twin pairs) from an existing twin-family registry based in Virginia, North Carolina, and South Carolina to (II) identify 16,000 individuals who are lifetime regular smokers (-62 percent current, -28 percent past) and obtain a comprehensive baseline set of predictors and correlates of smoking cessation and relapse; and (III) follow the smoking status of these lifetime regular smokers prospectively at yearly intervals for three years to identify smoking cessation and relapse events. These efforts will yield complimentary prospective and retrospective data sets. Analysis of these data sets will allow us: to describe changes in smoking behavior over time; to identify the predictors of progression in readiness to change, persistent reduction in amount smoked, serious attempts at smoking cessation, successful smoking cessation, and smoking relapse; to use multivariate classification techniques to derive a typology applicable to individual smokers that identifies more homogeneous subgroups; to model the smoking process (i.e., smoking initiation-regular smoking-smoking cessation OR smoking initiation-cessation-relapse) to determine the magnitude and overlap of the genetic and environmental sources of variation in the processes; and to investigate the co-variation of the etiological genetic and environmental sources of variation in these processes with measures of personality, self-esteem, self-efficacy, and psychopathology. Accomplishing the aims of this proposal will advance significantly our knowledge of the determinants of smoking cessation and relapse The use of a twin design, a large/powerful sample, multivariate assessment of the key phenotypes, and advanced analytic methods are key components of our design and approach.